Crazyflie coordinate system to ENU for Matlab

[jwdev297]

I need to work with the data of the Crazyflie 2.0 IMU. Therefore i created a log with the microSD adapter. I converted the data to a csv and then to mat-file.

When i now try to use Matlab Complementary Filter (https://de.mathworks.com/help/nav/ref/complementaryfilter-system-object.html) it always fails althrough I swapped x and y and changed the foresign of y also known as theta (to fit in general ENU instead of Crazyflie ENU).

I also tried just changing foresign of y without swapping axes.

....loading data to script as row vectors (with size 3x13744) and convert to general ENU. Then transpose the vectors (required for Matlab Complementary filter)...

Fs = 250 %Hz
refFrame = 'ENU'
decim = 1;

cfToENU = [0 1 0;
           -1 0 0;
           0 0 1];

for i=1:length(acc)
    acc(:,i) = cfToENU * acc(:,i);
    gyro(:,i) = cfToENU * gyro(:,i);
end
acc = acc.' * 9.80665;
gyro = gyro.';
time = time.';

fuse = complementaryFilter('SampleRate',Fs, 'ReferenceFrame',refFrame, 'HasMagnetometer', false)

q = fuse(acc, gyro);

time = (0:decim:size(acc,1)-1)/Fs;

plot(time,eulerd(q,'ZYX','frame'))
title('Orientation Estimate')
legend('Z-axis', 'Y-axis', 'X-axis')
xlabel('Time (s)')
ylabel('Rotation (degrees)')

Output of plot is the following:

Estimation makes no sense 😢

Matlab Complementary works with data not from the crazyflie (direct in general ENU).

What is the correct way in Matlab to do the convert for crazyflie data created with microsd adapter?
I really need to reach general ENU notation.

Thank you very much in advance

Script i used for convert from crazyflie binary to csv is the following:

# -*- coding: utf-8 -*-
"""
Helper to decode binary logged sensor data from crazyflie2 with uSD-Card-Deck
and save the decoded data to CSV files.

Usage:
    python3 usdlogToCSV.py  log07 --output_folder out/
"""
import sys
import os
import argparse
import csv

sys.path.append('../../tools/usdlog/')

from cfusdlog import decode

def save_to_csv(decoded_data, input_filename, output_folder="output"):
    """
    Save decoded log data to CSV files with the input filename as prefix.

    Parameters:
    - decoded_data: Dictionary containing decoded log data.
    - input_filename: The original input filename (used for prefixing).
    - output_folder: Directory where CSV files will be saved.
    """
    # Ensure output directory exists
    os.makedirs(output_folder, exist_ok=True)

    # Extract the base filename (without extension) to use as a prefix
    base_filename = os.path.splitext(os.path.basename(input_filename))[0]

    for event_name, event_data in decoded_data.items():
        # Set up the CSV file path with the prefix
        output_file = os.path.join(output_folder, f"{base_filename}_{event_name}.csv")
        
        try:
            # Open and write data to the CSV file
            with open(output_file, mode='w', newline='') as csvfile:
                writer = csv.writer(csvfile)

                # Write the header row
                headers = ['timestamp'] + list(event_data.keys())
                headers.remove('timestamp')  # Avoid duplicate timestamp column
                writer.writerow(headers)

                # Write rows of data
                rows = zip(event_data['timestamp'], *[event_data[k] for k in event_data if k != 'timestamp'])
                writer.writerows(rows)

            print(f"Saved {event_name} data to {output_file}")

        except Exception as e:
            print(f"Error saving {event_name} data to {output_file}: {e}")

if __name__ == "__main__":
    parser = argparse.ArgumentParser()
    parser.add_argument("filename", help="Path to the binary log file")
    parser.add_argument("--output_folder", default="output", help="Folder to save CSV files")
    args = parser.parse_args()

    try:
        # Decode the log file
        log_data = decode(args.filename)

        if log_data:
            # Save the decoded data to CSV files
            save_to_csv(log_data, args.filename, args.output_folder)
        else:
            print("No data found in the log file to save.")
    except FileNotFoundError:
        print(f"File not found: {args.filename}")
    except Exception as e:
        print(f"An error occurred while processing the log file: {e}")

The correct plot done with crazyflie example.py looks exactly:

[gemenerik]

At Bitcraze, we don’t actively use MATLAB ourselves, so we might not be the best to help with MATLAB-specific issues. However, what we can help with is understanding the Crazyflie coordinate system and making sure your transformation is correct.

Carefully going over the coordinate transformations and ensuring correct conversions would be a good place to start.

That said, others in the community may have worked with MATLAB and Crazyflie data before. It might be worth reaching out to see if someone with MATLAB experience can share their insights. Let us know if you have specific questions about the Crazyflie’s coordinate system—we’re happy to help with that!

[gemenerik]

https://github.com/search?q=Crazyflie+language%3AMATLAB&type=repositories&l=MATLAB

[jwdev297]

Thank you for your response!

While research for repositories was also helpful, I noticed that I load my gyro data in the wrong unit (deg/s instead of rad/s).